Music Selection Method Appropriate for Users' Situations and its Application Using Kansei Evaluation Only by the Existence of Impression

Abstract

A music selection method appropriate for users' various situations is proposed in this paper by accumulating only information whether each music has specific impression or not. For its application, a method which predicts users' situations and create appropriate playlists automatically from their music selection is also proposed. Since the validity of the above two points were confirmed, an easy kansei evaluation by the existence of impression is very useful for the music recommendation.